Upload
others
View
2
Download
0
Embed Size (px)
Citation preview
University of Nebraska - LincolnDigitalCommons@University of Nebraska - Lincoln
Bureau of Business Research Publications Bureau of Business Research
11-2013
The Role of Parental Experience inEntrepreneurship Choice among AdultsEric ThompsonUniversity of Nebraska-Lincoln, [email protected]
Carlos AsartaUniversity of Nebraska-Lincoln, [email protected]
Ziwen ZhangUniversity of Wisconsin Stevens Point, [email protected]
Felipe LeMarieUniversity of Nebraska-Lincoln, [email protected]
Follow this and additional works at: http://digitalcommons.unl.edu/bbrpub
This Article is brought to you for free and open access by the Bureau of Business Research at DigitalCommons@University of Nebraska - Lincoln. It hasbeen accepted for inclusion in Bureau of Business Research Publications by an authorized administrator of DigitalCommons@University of Nebraska -Lincoln.
Thompson, Eric; Asarta, Carlos; Zhang, Ziwen; and LeMarie, Felipe, "The Role of Parental Experience in Entrepreneurship Choiceamong Adults" (2013). Bureau of Business Research Publications. 40.http://digitalcommons.unl.edu/bbrpub/40
0
The Role of Parental Experience in
Entrepreneurship Choice among Adults
November 2013
Eric Thompson1, Carlos Asarta2, Ziwen Zhang3, and Felipe LeMarie4
Presented at the 2013 North American Meetings of the Regional Science Association International
Abstract – Entrepreneurship choice is influenced by a number of contemporaneous variables including
household wealth, spouse's education, gender, family composition, and health. Family background is
another important factor; in particular whether a parent was involved in entrepreneurship. The long-
running Panel Study of Income Dynamics provides an opportunity to examine the influence of parental
entrepreneurship, from the 1960s and 1970s, on the entrepreneurship choices of middle-age Americans
in the late 2000s. Consistent with previous literature, we find that parental experience is a consistent
predictor of entrepreneurship choice for contemporary Americans. Our results, however, shed light in
particular on the contribution of parental entrepreneurship when an individual was of school age or a
young adult.
1 Eric Thompson, Associate Professor, Department of Economics, University of Nebraska-Lincoln,
[email protected] 2 Carlos Asarta, Associate Professor, Department of Economics, University of Delaware, [email protected] 3 Ziwen Zhang, Assistant Professor, Department of Economics, University of Wisconsin Stevens Point,
[email protected] 4 Filipe LeMarie, Graduate Student, Department of Economics, University of Nebraska-Lincoln,
1
Introduction
The choice to pursue entrepreneurship is a complex decision influenced by entrepreneurial skill, tolerance
for risk, the amount of risk involved, and the resources available to potential entrepreneurs. Among these
variables, there are multiple measures of the resources available to entrepreneurs including wealth or
other sources of family income (Evans and Jovanovic, 1989; Gurley-Calvez, Hammond and Thompson,
2010). More mercurial are the factors that lead to risk tolerance and underlying entrepreneurial skill.
These traits relate to the idiosyncratic characteristics of individual personalities (Hmielski and Corbett,
2008; Chen, Greene and Crick, 1998) – factors revealed more through personality tests than through the
socio-economic characteristics of individuals and households. The potential presence of these
characteristics, however, can be related through simple observation for one group of individuals – the
children of entrepreneurs. These children may have both inherited and learned a taste for the risks and
rewards of entrepreneurship while at the same time learning the skills of entrepreneurs directly from their
parents.
At very least the children of entrepreneurs have multiple avenues with which to arrive at a career in
entrepreneurship. And, turning the issue around, an individual’s choice of a career in entrepreneurship
may be influenced by whether their parents were entrepreneurs. This paper examines the choice of
individuals to engage in entrepreneurship over a recent three year period, from 2005 to 2007, controlling
for a variety of other factors which influence entrepreneurship choice such as personal characteristics and
household resources. The study utilizes data from the Panel Study of Income Dynamics to gain detailed
information about parents and their labor market activities over a period of decades, allowing our model
to include variables about the specific timing of parent entrepreneurship.
The next section of the paper examines literature on entrepreneurship choice to identify factors which
contribute to the choice of entrepreneurship. The third section describes the data that is available
through the Panel Study of Income Dynamics. The fourth section contains regression results and the fifth
section is the conclusion.
2
Literature Review
This section examines both individual and household characteristics that are associated with
entrepreneurship and the propensity to choose a career in entrepreneurship. Key individual
characteristics include gender, education, and health status. Among these, gender may influence the
probability of entrepreneurship given that there are more male sole-proprietors than female sole-
proprietors. This said, many businesses are jointly owned by both male and female entrepreneurs.
Further, the gap between male and female entrepreneurship rates has been in decline. Lowrey (2005),
using individual tax returns from IRS Statistics of Income data, found that between 1985 and 2000, the
number of female sole-proprietors grew by 82%, compared to 39% growth in male sole-proprietorships.
The trend continued in the years that followed. Gurley-Calvez, Hammond and Thompson (2010) found
that the share of women engaged in self-employment rose faster than the share of men during the 2000
to 2006 period in U.S. labor market areas. This suggests gender may not have a significant influence on
entrepreneurship choice during the period we will examine in the late 2000s. Education is also positively
associated with the probability of choosing a career in entrepreneurship. In particular Weaver, Dickson
and Solomon (2006) and Moultry (2007) find that the probability of selecting self-employment over wage
and salary employment rises with education attainment.
Household characteristics also influence the probability of choosing a career in entrepreneurship. Key
households characteristics include the availability of time and resources in terms of spouse income and
wealth. Moutray (2007), utilizing data from the Panel Study of Income Dynamics, finds that married
individuals are more likely to be self-employed. Further, wealth is thought to be positively correlated with
entrepreneurship (Evans and Jovanovic, 1989). This could occur in part due to liquidity constraints, where
entrepreneurs must provide a substantial share of the capital to start their business, as self-selection and
moral hazard limit the ability to borrow money to start a business. Indeed, Dunn and Holtz-Eakin (1996)
find that the probability of transition into entrepreneurship increases both with an individual’s wealth but
also with their parent’s wealth. Evans and Jovanovic (1989) further argue that wealth may be correlated
with latent entrepreneurial ability and may make individuals less risk averse, and therefore, better able to
3
engage in entrepreneurship. Another possibility is that individuals recognize their own entrepreneurial
potential and individuals are more likely to save and accumulate wealth early in life for the purpose of
starting a business. Xu (1998) focuses on the latter argument and, unlike Evans and Jovanovic, finds a
correlation between wealth and entrepreneurial ability.
Parental entrepreneurship is another key characteristic influencing whether individuals choose a career in
self-employment. This can be thought of as an extended family characteristic. The children of
entrepreneurs may have inherited or learned risk-tolerance and entrepreneurial skills from their parents
that increase their probability of choosing a career in entrepreneurship. Lindquist, Sol and Van Praag
(2012), studying a data set of Swedish adoptees, find significantly higher rates of self-employment as an
adult among children who were adopted by self-employed parent, and among children whose birth
parents were self-employed. The effect of self-employed adoptive parents (nurture) was twice that of
self-employed birth parents (nature). Dunn and Holtz-Eakin (1996), using data from the National
Longitudinal Surveys, find that parent entrepreneurship had a positive and statistically significant
influence on the probability that an individual would transition from wage and salary employment to self-
employment in a given year. Having a parent or parents who were self-employed doubled the probability
that an individual would transition from wage and salary work to self-employment.
Environmental factors also influence the probability of engaging in entrepreneurship. Greater information
about the feasibility of entrepreneurship within a localized area can improve its likelihood, both by
reducing risk for potential entrepreneurs and reducing the cost of obtaining financial services such as
loans or insurance, as argued in Bunten, et al. (2013). Unfortunately, our research team has not yet
gained access to the geo-coded micro-sample of the Panel Study of Income Dynamics in order to obtain
information about the county of residence for individuals within the sample.
4
Panel Study of Income Dynamics Data
The Panel Study of Income Dynamics (PSID), which is maintained by the Institute for Social Research at
the University of Michigan, follows households and the individuals within those households over an
extended period beginning in 1968. The survey began with an original sample of 5,000 households and
followed those households annually over the next four and one-half decades. In each annual or biennial
survey (surveys were biennial beginning in 1997), information is gathered about all individuals in a
household but the most information is gathered about the head of the household and their spouse.
Information is gathered on a variety of economic and demographic variables, including information about
work, household wealth, health and family structure. Importantly, the Panel Study of Income Dynamics
follows new households when these households are formed by members of existing households within
the sample. In particular, children in the original surveyed households split off to form their own
households which are then tracked along with households from the original sample. This makes it
possible to follow children from the original sample of households and learn how their work lives
progressed, and relate those developments to the characteristics of the households in which they were
raised. Our goal is to evaluate whether the children of Panel Study households choose to engage in
entrepreneurship as adults, and whether that choice was influenced by the career choices of their
parents.
From the original households in the Panel Study of Income Dynamics, we gathered data on all individuals
who turned 18 years of age from 1976 to 1979. This allowed us to observe the self-employment behavior
of individuals while having detailed information about the employment status of their parents, directly
reported by the parents in earlier panels of the PSID. There were a total of 620 men and women who
turned 18 during that 4-year period. We then checked for the self-employment status of these individuals’
decades later, during the period 2005 through 2007. Individuals were considered self-employed if they
were self-employed at some time during that 3-year period. Three years were used in order to capture
individuals who spent both some years self-employed and some years engaged in wage and salary work,
as well as individuals who were self-employed throughout the three-year period. Analysis stopped in 2007
5
in order to avoid self-employment behavior which was influenced by the “Great Recession” that began in
December 2007. Self-employment might be artificially elevated during recession years, out of necessity.
Table 1 shows the un-weighted mean values and standard deviations for all key variables. Of the 620
individuals in our sample, 165 (26.5%) were self-employed between 2005 and 2007.
Table 1 also includes sample statistics for a set of individual and household characteristics that influence
the choice of a career in entrepreneurship. Personal characteristics include gender and health. The
majority of individuals in the sample were female (55.2%). Health status could influence the choice of
entrepreneurship given the rigors and time commitment associated with running a business. Healthier
individuals may be better prepared for these responsibilities. The Panel Study of Income Dynamics
included questions about the general health of individuals included in the survey during its 2007
interviews. Self-assessed health was reported in the following five categories: excellent, very good, good,
fair, and poor. Even the collection of category titles reflects the positive tendencies in self-assessment of
health, with most categories reflecting gradations of good health. Indeed, a majority of the sample
reports itself to be in excellent (16.5%) or very good health (35.0%) with another 31.5% reporting good
health. Only about one in six covered individuals in the survey reported themselves to be in fair (13.2%)
or poor (3.5%) general health. In the regression analysis, health status variables are entered into the
model as a set of binary variables.
Turning to household characteristics, variables were included on household characteristics that could
influence the resources available to would be entrepreneurs. Marital status was one measure of
household resources, following Moutray (2007). More than half of the sample was married (54.5%) while
23.2% were divorced, 5.5% were separated and 15.0% were never married. A handful of individuals
were widows or widowers. A set of binary variables are used in the regression analysis to evaluate the
influence of marital status on the choice of a career in entrepreneurship.
Along the same lines, the income potential of the spouse of married individuals also could influence
entrepreneurship choice. We include a variable for spouse education for the more than 330 individuals in
the sample who are married, given the correlation between education and earnings. The average years of
6
education among spouses in 2007 was 13.44 years, reflecting the fact that most individuals in their 30s,
40s, and 50s (who would be married to our sample of persons who turned 18 during the 1976 to 1979
period) have participated in post-secondary education at some point in their lives. A variable for
household assets also was included in the model. Among individuals in the sample, the average value of
household assets was approximately $19,800 in 2007, though there was substantial variability in asset
values, as can be seen in Table 1.
We also utilized the long family histories available in the PSID to gather information about parent
entrepreneurship. Moving through the years of the panel, we examined the employment status of the
parents of our 620 18-year olds in each year. This information was used to create variables indicating
whether parents were self-employed during the childhood (age 0-17), young adulthood (age 18-22)
and/or adulthood of each individual, as seen in Table 1. The Parent Entrepreneur Early variable takes a
value of 1 if the mother or father of an individual was ever self-employed when that individual was
between the ages of 0 and 17. It takes a value of 0 if not. For 130 individuals, at least one parent was
self-employed for at least one year while the individual was age 0 to 17. For 113 individuals, at least one
parent was self-employed during at least one year when the individual was age 18 to 22. For these
individuals, the Parent Entrepreneurship Middle variable received a value of 1. Finally, individuals received
a value of 1 for the Parent Entrepreneurship Late variable if at least one parent was self-employed at
some point when the individual was age 23 or older. These three variables are utilized in our model of
entrepreneurship choice in order to isolate the role of parent entrepreneurship when individuals are
school-age (0-17) or young adults (18-22). In particular, by deploying all three variables in a model, we
can isolate the impact of parent entrepreneurship in childhood or young adulthood, after controlling for
whether a parent was also an entrepreneur when an individual was an adult. This isolates the learning
aspects of parent entrepreneurship rather than the succession aspects of parent entrepreneurship, such
as business ownership transitions.
7
Table 1
Sample Statistics
Variable
N
Mean
Standard
Deviation
Minimum
Value
Maximum
Value
Dependent Variable
Self-Employed 2005-2007 (Yes=1) 620 0.266 0.442 0 1
Independent Variables
Female (Yes=1) 620 0.552 0.498 0 1
Married (Yes=1) 619 0.545 0.498 0 1
Never Married (Yes=1) 619 0.150 0.357 0 1
Widow/Widower (Yes=1) 619 0.016 0.126 0 1
Divorced (Yes=1) 619 0.232 0.423 0 1
Separated (Yes=1) 619 0.055 0.228 0 1
Excellent General Health (Yes=1) 618 0.165 0.371 0 1
Very Good General Health (Yes=1) 618 0.350 0.477 0 1
Good General Health (Yes=1) 618 0.315 0.465 0 1
Fair General Health (Yes=1) 618 0.132 0.339 0 1
Poor General Health (Yes=1) 618 0.035 0.185 0 1
Years of Education Spouse 331 13.444 2.077 8 17
Value of Assets 567 $19,787 $82,950 $0 $1,500,000
Parent Entrepreneur Early (Yes=1) 620 0.210 0.407 0 1
Parent Entrepreneur Middle (Yes=1) 620 0.182 0.386 0 1
Parent Entrepreneur Late (Yes=1) 620 0.298 0.458 0 1
Source: Individuals drawn from the Panel Study of Income Dynamics
8
Results
A logit model was used to estimate the likelihood that individuals who turned 18 between 1976 and 1979
would choose a career in entrepreneurship (i.e., self-employment) during adulthood, specifically during
the years 2005 and 2007. Following the literature, entrepreneurship was defined as self-employment.
Coefficient estimates from the logit model are reported in Table 2. Marginal probabilities are not
calculated. However, the sign of the coefficient estimates would be the same as that for the marginal
probabilities. In any case, the direction of the relationship between independent variables and the
probability of entrepreneurship can be gleaned from results in Table 2.
In the model, the choice of self-employment was a function of personal characteristics such as gender
and health. The choice of self-employment was also a function of household characteristics such as
marital status, value of assets or the education of a spouse, if one was present. The last three variables
were related to parent entrepreneurship during childhood (ages 0-17), young adulthood (ages 18-22) and
adulthood (ages 23 and above).
Results of the logit model are reported in Table 2. The first variable in the model, Female, takes a value
of 1 if the individual in the sample is female and a value of 0 if the individual is male. The coefficient
estimate for the Female variable is negative. However, the coefficient estimate is not statistically
significant. The probability of choosing a career of entrepreneurship was not found to be related to
gender.
We next examine the influence of two key personal and family characteristics that could influence
entrepreneurship as a career choice: marital status and health status. A set of binary variables are
presented for marital status. Married is the omitted marital status category. As seen in Table 2, there was
no statistically significant difference in the probability of entrepreneurship between married individuals
and individuals with a different marital status. While the point estimate for the coefficient for the never
married status was positive and would be statistically significant under a 15% confidence level, results in
general indicate that marital status did not have a significant influence on the choice of a career in
9
entrepreneurship. This result is in contrast to Moutray (2007). Binary variables for health status provide a
continuum for general health with excellent health as the omitted category. Given the rigors of business
ownership, we anticipated to see the probability of entrepreneurship fall with health status. However, we
found no difference in the probability of choosing a career in entrepreneurship between individuals who
describe themselves as being in excellent general health and those in good general health, fair general
health or poor general health. These results in general suggest that the choice of a career in
entrepreneurship is not related to health status. One interesting result is that individuals who describe
themselves as having very good general health have a higher probability of choosing entrepreneurship
than individuals who describe themselves as having excellent health. This result is difficult to explain
given that the two categories of excellent and very good health status are near to each other in terms of
health assessment. The difference in self-assessment may simply reflect the type of sober self-
assessment that entrepreneurs must make on a regular basis.
In general, we found that health and marital status did not have a consistent relationship with the
probability of a career in entrepreneurship. To examine the issue further, we also interact health and
marital status in the model reported in Table 2. However, there were no statistically significant results for
these interaction terms and the specific results are not reported in the table.
Additionally, we add a variable controlling for the education of the spouse of married individuals into the
model. In particular, we have a variable for the spouse’s years of education. Coefficient estimates on the
variable indicate whether the likelihood of choosing a career in entrepreneurship varies with the
education of the spouse. That likelihood would be expected to rise because a spouse’s education would
be correlated with a spouse’s potential earnings. Higher earnings by a spouse may create financial
resources that enable more individuals to engage in inherently risky entrepreneurship. An alternative
explanation is that there is a correlation between the education attainment of husbands and wives, and
spouse’s education is acting as a proxy for individual education. Results in Table 2 show that the
coefficient on spouse’s years of education is positive and statistically significant at the 5% confidence
level.
10
The model also contains another measure of family resources, in particular, the value of household
assets. Wealth, as noted by Evans and Jovanovic (1989) and Xu (1998), can be positively correlated with
entrepreneurship either because wealth helps address a liquidity constraint faced by entrepreneurs or is
an indicator of latent entrepreneurial ability. Within our model, when the wealth variable is measured
contemporaneous with entrepreneurship, wealth also may reflect the higher financial rewards earned by
entrepreneurs. As seen in Table 2, the coefficient on the value of family assets, in dollars, is positive and
statistically significant at the 10% confidence level.
Table 2 also contains three variables related to parent entrepreneurship. The variables are interpreted as
three sets of binary variables since the three categories are not mutually exclusive. An individual’s mother
and/or father may have been self-employed in two or even all three periods. Parent entrepreneurship in
all three periods was positively associated with the likelihood that an individual would choose a career as
an entrepreneur as an adult. In particular, parent entrepreneurship when an individual was school age
(0-17) or a young adult (18-22) had a positive influence on subsequent entrepreneurship choice as an
adult, even after controlling for whether an individual’s parent were entrepreneurs when that individual
was an adult. These results suggest a role for learning entrepreneurial skills in the household setting both
as a child and as a young adult.
11
Table 2
Regression Results: Factors Influencing Self-Employment 2005-2007
Variable Estimate (Standard Deviation in Parenthesis)
Intercept -1.52** (0.64)
Female -0.0041
(0.21)
Never Married 1.89
(1.31)
Widow/Widower -13.97 (102.9)
Divorced 0.42
(.66)
Separated -13.97
(102.9)
Very Good General Health 0.97** (0.38)
Good General Health 0.38
(.39)
Fair General Health 0.39
(.53)
Poor General Health 0.29 (1.21)
Years of Education Spouse 0.068**
(0.030)
Value of Assets ($) 5.99 x 106*
(3.09 x 106)
Parent Entrepreneur Early 0.63** (0.32)
Parent Entrepreneur Middle 1.54***
(0.38)
Parent Entrepreneur Late 0.65**
(0.26)
Interactions for Marital Status and Health Yes
N 567
*=Statistically significant at 10% level, **=statistically significant at 5% level, ***=statistically significant
at 1% level.
Conclusion
The choice to pursue a career in entrepreneurship is a complex decision influenced by personal and
household characteristics. Among the key factors are entrepreneurial skills and a tolerance of risk. These
factors are both inherent in individuals and learned. In particular, the entrepreneurial tendency of parents
12
is a key factor in both inherent and learned entrepreneurial skills and risk tolerance. This study examines
the role of parent entrepreneurship in both childhood and young adulthood on the subsequent
entrepreneurship of individuals as middle-age adults. The analysis relies on the detailed tracking of
families over generations in the Panel Study of Income Dynamics. In particular, the long-running Panel
Study provides an opportunity to examine the influence of parental entrepreneurship, from the 1960s and
1970s, on the entrepreneurship choices of current middle-age Americans in the late 2000s. We find that
parental experience is a consistent predictor of entrepreneurship choice for contemporary Americans. In
particular, parent entrepreneurship when an individual was school age or a young adult was positively
correlated with an individual’s subsequent choice of entrepreneurship as an adult. This result held after
adjusting for whether an individual’s parent where entrepreneurs when that individual was an adult.
13
References
Bunten, Devin, Stephen Weiler, Sammy Zahran, and Eric Thompson, 2013. “Entrepreneurship,
Information, and Growth.” Presented at the 60th Annual North American Meetings of the
Regional Science Association International, Atlanta, Georgia (November).
Chen, Chao, Patricia Greene and Ann Crick, 1998. “Does Entrepreneurial Self-Efficacy Distinguish
Entrepreneurs From Managers?” Journal of Business Venturing, 13: 295-316.
Dunn, Thomas, and Douglas Holtz-Eakin, 1996. “Financial Capital, Human Capital, and the Transition to
Self-Employment: Evidence from Intergenerational Linkds,” NBER Working Paper 5622 (June).
Evans, David and Boyan Jovanovic, 1989. “An Estimated Model of Entrepreneurial Choice under Liquidity
Constraints,” Journal of Political Economy, 97(4): 808-827.
Gurley-Calvez, Tami, George Hammond and Eric Thompson. 2010. “Determinants of Entrepreneurship
across U.S. Labr Market Areas, 1970 to 2006.” Research Report. U.S. Small Business
Administration, Office of Advocacy. February 2010.
Hmielski, Keith and Andrew Corbett, 2008. “The contrasting interaction effects of improvisational
behavior with entrepreneurial self-efficacy on new venture performance and entrepreneur work
satisfaction,” Journal of Business Venturing, 23: 482-496.
Lindquist, Matthew, Joeri Sol, and Mirjam Van Praag, 2012. “Do Entrepreneurial Parents Have
Entrepreneurial Children?,” IZA Discussion Paper No. 6740 (September)
Lowrey, Ying, 2005. “U.S. Sole-Proprietorships: A Gender Comparison, 1985-2000.” Research Report.
U.S. Small Business Administration, Office of Advocacy. September 2005.
Moutray, Chad. 2007. “Educational Attainment and Other Characteristics of the Self-Employed: An
Examination Using Data from the Panel Study of Income Dynamics.” Research Report. U.S. Small
Business Administration, Office of Advocacy. December 2007.
14
Weaver, Mark, Pat Dickson, and George Solomon. 2006. “Entrepreneurship and Education: What is
Known and Not Known about the Links Between Education and Entrepreneurial Activity.” The
Small Business Economy: A Report to the President, 2006. Chapter 5. Washington, D.C.: U.S.
Small Business Administration, Office of Advocacy. December 2006.
Xu, Bin, 1998. “A re-estimation of the Evans – Jovanovic entrepreneurial choice model,” Economics
Letters 58: 91-95.